Skip to content

Commit

Permalink
Delete shard store files before restoring a snapshot (#27476)
Browse files Browse the repository at this point in the history
Pull request #20220 added a change where the store files
that have the same name but are different from the ones in the
snapshot are deleted first before the snapshot is restored.
This logic was based on the `Store.RecoveryDiff.different`
set of files which works by computing a diff between an
existing store and a snapshot.

This works well when the files on the filesystem form valid
shard store, ie there's a `segments` file and store files
are not corrupted. Otherwise, the existing store's snapshot
metadata cannot be read (using Store#snapshotStoreMetadata())
and an exception is thrown
(CorruptIndexException, IndexFormatTooOldException etc) which
is later caught as the begining of the restore process
(see RestoreContext#restore()) and is translated into
an empty store metadata (Store.MetadataSnapshot.EMPTY).

This will make the deletion of different files introduced
in #20220 useless as the set of files will always be empty
even when store files exist on the filesystem. And if some
files are present within the store directory, then restoring
a snapshot with files with same names will fail with a
FileAlreadyExistException.

This is part of the #26865 issue.

There are various cases were some files could exist in the
 store directory before a snapshot is restored. One that
Igor identified is a restore attempt that failed on a node
and only first files were restored, then the shard is allocated
again to the same node and the restore starts again (but fails
 because of existing files). Another one is when some files
of a closed index are corrupted / deleted and the index is
restored.

This commit adds a test that uses the infrastructure provided
by IndexShardTestCase in order to test that restoring a shard
succeed even when files with same names exist on filesystem.

Related to #26865
  • Loading branch information
tlrx committed Nov 24, 2017
1 parent 3bc3d4d commit 506449b
Show file tree
Hide file tree
Showing 5 changed files with 209 additions and 18 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -731,7 +731,7 @@ public String toString() {

/**
* Represents a snapshot of the current directory build from the latest Lucene commit.
* Only files that are part of the last commit are considered in this datastrucutre.
* Only files that are part of the last commit are considered in this datastructure.
* For backwards compatibility the snapshot might include legacy checksums that
* are derived from a dedicated checksum file written by older elasticsearch version pre 1.3
* <p>
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,6 @@
import org.apache.lucene.store.RateLimiter;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.BytesRefBuilder;
import org.apache.lucene.util.IOUtils;
import org.elasticsearch.ElasticsearchParseException;
import org.elasticsearch.ExceptionsHelper;
import org.elasticsearch.ResourceNotFoundException;
Expand Down Expand Up @@ -110,6 +109,7 @@
import java.nio.file.FileAlreadyExistsException;
import java.nio.file.NoSuchFileException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.Collections;
import java.util.HashMap;
Expand Down Expand Up @@ -1451,6 +1451,9 @@ public void restore() throws IOException {
SnapshotFiles snapshotFiles = new SnapshotFiles(snapshot.snapshot(), snapshot.indexFiles());
Store.MetadataSnapshot recoveryTargetMetadata;
try {
// this will throw an IOException if the store has no segments infos file. The
// store can still have existing files but they will be deleted just before being
// restored.
recoveryTargetMetadata = targetShard.snapshotStoreMetadata();
} catch (IndexNotFoundException e) {
// happens when restore to an empty shard, not a big deal
Expand Down Expand Up @@ -1478,7 +1481,14 @@ public void restore() throws IOException {
snapshotMetaData.put(fileInfo.metadata().name(), fileInfo.metadata());
fileInfos.put(fileInfo.metadata().name(), fileInfo);
}

final Store.MetadataSnapshot sourceMetaData = new Store.MetadataSnapshot(unmodifiableMap(snapshotMetaData), emptyMap(), 0);

final StoreFileMetaData restoredSegmentsFile = sourceMetaData.getSegmentsFile();
if (restoredSegmentsFile == null) {
throw new IndexShardRestoreFailedException(shardId, "Snapshot has no segments file");
}

final Store.RecoveryDiff diff = sourceMetaData.recoveryDiff(recoveryTargetMetadata);
for (StoreFileMetaData md : diff.identical) {
BlobStoreIndexShardSnapshot.FileInfo fileInfo = fileInfos.get(md.name());
Expand All @@ -1505,29 +1515,31 @@ public void restore() throws IOException {
logger.trace("no files to recover, all exists within the local store");
}

if (logger.isTraceEnabled()) {
logger.trace("[{}] [{}] recovering_files [{}] with total_size [{}], reusing_files [{}] with reused_size [{}]", shardId, snapshotId,
index.totalRecoverFiles(), new ByteSizeValue(index.totalRecoverBytes()), index.reusedFileCount(), new ByteSizeValue(index.reusedFileCount()));
}
try {
// first, delete pre-existing files in the store that have the same name but are
// different (i.e. different length/checksum) from those being restored in the snapshot
for (final StoreFileMetaData storeFileMetaData : diff.different) {
IOUtils.deleteFiles(store.directory(), storeFileMetaData.name());
}
// list of all existing store files
final List<String> deleteIfExistFiles = Arrays.asList(store.directory().listAll());

// restore the files from the snapshot to the Lucene store
for (final BlobStoreIndexShardSnapshot.FileInfo fileToRecover : filesToRecover) {
// if a file with a same physical name already exist in the store we need to delete it
// before restoring it from the snapshot. We could be lenient and try to reuse the existing
// store files (and compare their names/length/checksum again with the snapshot files) but to
// avoid extra complexity we simply delete them and restore them again like StoreRecovery
// does with dangling indices. Any existing store file that is not restored from the snapshot
// will be clean up by RecoveryTarget.cleanFiles().
final String physicalName = fileToRecover.physicalName();
if (deleteIfExistFiles.contains(physicalName)) {
logger.trace("[{}] [{}] deleting pre-existing file [{}]", shardId, snapshotId, physicalName);
store.directory().deleteFile(physicalName);
}

logger.trace("[{}] [{}] restoring file [{}]", shardId, snapshotId, fileToRecover.name());
restoreFile(fileToRecover, store);
}
} catch (IOException ex) {
throw new IndexShardRestoreFailedException(shardId, "Failed to recover index", ex);
}
final StoreFileMetaData restoredSegmentsFile = sourceMetaData.getSegmentsFile();
if (recoveryTargetMetadata == null) {
throw new IndexShardRestoreFailedException(shardId, "Snapshot has no segments file");
}
assert restoredSegmentsFile != null;

// read the snapshot data persisted
final SegmentInfos segmentCommitInfos;
try {
Expand Down Expand Up @@ -1602,5 +1614,4 @@ private void restoreFile(final BlobStoreIndexShardSnapshot.FileInfo fileInfo, fi
}
}
}

}
Original file line number Diff line number Diff line change
Expand Up @@ -188,7 +188,7 @@ public void restoreSnapshot(final RestoreRequest request, final ActionListener<R
final SnapshotInfo snapshotInfo = repository.getSnapshotInfo(snapshotId);
final Snapshot snapshot = new Snapshot(request.repositoryName, snapshotId);
List<String> filteredIndices = SnapshotUtils.filterIndices(snapshotInfo.indices(), request.indices(), request.indicesOptions());
MetaData metaData = repository.getSnapshotMetaData(snapshotInfo, repositoryData.resolveIndices(filteredIndices));
final MetaData metaData = repository.getSnapshotMetaData(snapshotInfo, repositoryData.resolveIndices(filteredIndices));

// Make sure that we can restore from this snapshot
validateSnapshotRestorable(request.repositoryName, snapshotInfo);
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,142 @@
/*
* Licensed to Elasticsearch under one or more contributor
* license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright
* ownership. Elasticsearch licenses this file to you under
* the Apache License, Version 2.0 (the "License"); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/

package org.elasticsearch.repositories.blobstore;

import org.apache.lucene.store.Directory;
import org.apache.lucene.util.IOUtils;
import org.apache.lucene.util.TestUtil;
import org.elasticsearch.cluster.metadata.RepositoryMetaData;
import org.elasticsearch.cluster.routing.ShardRouting;
import org.elasticsearch.cluster.routing.ShardRoutingHelper;
import org.elasticsearch.common.UUIDs;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.env.Environment;
import org.elasticsearch.index.shard.IndexShard;
import org.elasticsearch.index.shard.IndexShardState;
import org.elasticsearch.index.shard.IndexShardTestCase;
import org.elasticsearch.index.shard.ShardId;
import org.elasticsearch.index.store.Store;
import org.elasticsearch.index.store.StoreFileMetaData;
import org.elasticsearch.repositories.IndexId;
import org.elasticsearch.repositories.Repository;
import org.elasticsearch.repositories.fs.FsRepository;
import org.elasticsearch.snapshots.Snapshot;
import org.elasticsearch.snapshots.SnapshotId;

import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.Arrays;
import java.util.List;

import static org.elasticsearch.cluster.routing.RecoverySource.StoreRecoverySource.EXISTING_STORE_INSTANCE;

/**
* This class tests the behavior of {@link BlobStoreRepository} when it
* restores a shard from a snapshot but some files with same names already
* exist on disc.
*/
public class BlobStoreRepositoryRestoreTests extends IndexShardTestCase {

/**
* Restoring a snapshot that contains multiple files must succeed even when
* some files already exist in the shard's store.
*/
public void testRestoreSnapshotWithExistingFiles() throws IOException {
final IndexId indexId = new IndexId(randomAlphaOfLength(10), UUIDs.randomBase64UUID());
final ShardId shardId = new ShardId(indexId.getName(), indexId.getId(), 0);

IndexShard shard = newShard(shardId, true);
try {
// index documents in the shards
final int numDocs = scaledRandomIntBetween(1, 500);
recoverShardFromStore(shard);
for (int i = 0; i < numDocs; i++) {
indexDoc(shard, "doc", Integer.toString(i));
if (rarely()) {
flushShard(shard, false);
}
}
assertDocCount(shard, numDocs);

// snapshot the shard
final Repository repository = createRepository();
final Snapshot snapshot = new Snapshot(repository.getMetadata().name(), new SnapshotId(randomAlphaOfLength(10), "_uuid"));
snapshotShard(shard, snapshot, repository);

// capture current store files
final Store.MetadataSnapshot storeFiles = shard.snapshotStoreMetadata();
assertFalse(storeFiles.asMap().isEmpty());

// close the shard
closeShards(shard);

// delete some random files in the store
List<String> deletedFiles = randomSubsetOf(randomIntBetween(1, storeFiles.size() - 1), storeFiles.asMap().keySet());
for (String deletedFile : deletedFiles) {
Files.delete(shard.shardPath().resolveIndex().resolve(deletedFile));
}

// build a new shard using the same store directory as the closed shard
ShardRouting shardRouting = ShardRoutingHelper.initWithSameId(shard.routingEntry(), EXISTING_STORE_INSTANCE);
shard = newShard(shardRouting, shard.shardPath(), shard.indexSettings().getIndexMetaData(), null, null, () -> {});

// restore the shard
recoverShardFromSnapshot(shard, snapshot, repository);

// check that the shard is not corrupted
TestUtil.checkIndex(shard.store().directory());

// check that all files have been restored
final Directory directory = shard.store().directory();
final List<String> directoryFiles = Arrays.asList(directory.listAll());

for (StoreFileMetaData storeFile : storeFiles) {
String fileName = storeFile.name();
assertTrue("File [" + fileName + "] does not exist in store directory", directoryFiles.contains(fileName));
assertEquals(storeFile.length(), shard.store().directory().fileLength(fileName));
}
} finally {
if (shard != null && shard.state() != IndexShardState.CLOSED) {
try {
shard.close("test", false);
} finally {
IOUtils.close(shard.store());
}
}
}
}

/** Create a {@link Repository} with a random name **/
private Repository createRepository() throws IOException {
Settings settings = Settings.builder().put("location", randomAlphaOfLength(10)).build();
RepositoryMetaData repositoryMetaData = new RepositoryMetaData(randomAlphaOfLength(10), FsRepository.TYPE, settings);
return new FsRepository(repositoryMetaData, createEnvironment(), xContentRegistry());
}

/** Create a {@link Environment} with random path.home and path.repo **/
private Environment createEnvironment() {
Path home = createTempDir();
return new Environment(Settings.builder()
.put(Environment.PATH_HOME_SETTING.getKey(), home.toAbsolutePath())
.put(Environment.PATH_REPO_SETTING.getKey(), home.resolve("repo").toAbsolutePath())
.build());
}
}
Original file line number Diff line number Diff line change
Expand Up @@ -46,6 +46,7 @@
import org.elasticsearch.common.util.BigArrays;
import org.elasticsearch.common.xcontent.XContentType;
import org.elasticsearch.env.NodeEnvironment;
import org.elasticsearch.index.Index;
import org.elasticsearch.index.IndexSettings;
import org.elasticsearch.index.MapperTestUtils;
import org.elasticsearch.index.VersionType;
Expand All @@ -60,6 +61,7 @@
import org.elasticsearch.index.mapper.Uid;
import org.elasticsearch.index.seqno.SequenceNumbers;
import org.elasticsearch.index.similarity.SimilarityService;
import org.elasticsearch.index.snapshots.IndexShardSnapshotStatus;
import org.elasticsearch.index.store.DirectoryService;
import org.elasticsearch.index.store.Store;
import org.elasticsearch.indices.recovery.PeerRecoveryTargetService;
Expand All @@ -69,6 +71,9 @@
import org.elasticsearch.indices.recovery.RecoveryTarget;
import org.elasticsearch.indices.recovery.StartRecoveryRequest;
import org.elasticsearch.node.Node;
import org.elasticsearch.repositories.IndexId;
import org.elasticsearch.repositories.Repository;
import org.elasticsearch.snapshots.Snapshot;
import org.elasticsearch.test.DummyShardLock;
import org.elasticsearch.test.ESTestCase;
import org.elasticsearch.threadpool.TestThreadPool;
Expand All @@ -85,6 +90,7 @@
import java.util.function.BiFunction;
import java.util.function.Consumer;

import static org.elasticsearch.cluster.routing.TestShardRouting.newShardRouting;
import static org.hamcrest.Matchers.contains;
import static org.hamcrest.Matchers.hasSize;

Expand Down Expand Up @@ -583,6 +589,38 @@ protected void flushShard(IndexShard shard, boolean force) {
shard.flush(new FlushRequest(shard.shardId().getIndexName()).force(force));
}

/** Recover a shard from a snapshot using a given repository **/
protected void recoverShardFromSnapshot(final IndexShard shard,
final Snapshot snapshot,
final Repository repository) throws IOException {
final Version version = Version.CURRENT;
final ShardId shardId = shard.shardId();
final String index = shardId.getIndexName();
final IndexId indexId = new IndexId(shardId.getIndex().getName(), shardId.getIndex().getUUID());
final DiscoveryNode node = getFakeDiscoNode(shard.routingEntry().currentNodeId());
final RecoverySource.SnapshotRecoverySource recoverySource = new RecoverySource.SnapshotRecoverySource(snapshot, version, index);
final ShardRouting shardRouting = newShardRouting(shardId, node.getId(), true, ShardRoutingState.INITIALIZING, recoverySource);

shard.markAsRecovering("from snapshot", new RecoveryState(shardRouting, node, null));
repository.restoreShard(shard, snapshot.getSnapshotId(), version, indexId, shard.shardId(), shard.recoveryState());
}

/** Snapshot a shard using a given repository **/
protected void snapshotShard(final IndexShard shard,
final Snapshot snapshot,
final Repository repository) throws IOException {
final IndexShardSnapshotStatus snapshotStatus = new IndexShardSnapshotStatus();
try (Engine.IndexCommitRef indexCommitRef = shard.acquireIndexCommit(true)) {
Index index = shard.shardId().getIndex();
IndexId indexId = new IndexId(index.getName(), index.getUUID());

repository.snapshotShard(shard, snapshot.getSnapshotId(), indexId, indexCommitRef.getIndexCommit(), snapshotStatus);
}
assertEquals(IndexShardSnapshotStatus.Stage.DONE, snapshotStatus.stage());
assertEquals(shard.snapshotStoreMetadata().size(), snapshotStatus.numberOfFiles());
assertNull(snapshotStatus.failure());
}

/**
* Helper method to access (package-protected) engine from tests
*/
Expand Down

0 comments on commit 506449b

Please sign in to comment.